Brown vs Miliband

Both ICM and YouGov asked people about Brown and Miliband in their recent polls. In YouGov’s poll 21% of people thought Miliband would be a better Prime Minister than Brown (unfortunately YouGov didn’t give the alternative of him being worse than Brown – only of not being better). 11% of people told YouGov they would be more likely to vote Labour with Miliband as leader, 9% less likely.

ICM asked people to pick the best PM from Brown/Cameron and from Miliband/Cameron. In both cases Cameron lead by 21 points. 18% of respondents were more likely to vote Labour with Miliband as leader, 14% more likely to vote Labour with Brown. ICM also asked people to compare Brown and Miliband on various measures. Brown outranked Miliband in terms of honesty, making a stand on difficult isues and being a competent manager. He was also seen as more substantial and less likely to spin than Miliband. Miliband outranked Brown in terms of being “on my wavelength”, looking to the future and having the widest appeal. Whether this is particularly meaningful is a different question – how many of us really have any idea how competent a manager David Miliband is?

As Mike Smithson over on Political Betting has been saying these more likely/less likely questions aren’t particularly enlightening. Slightly better would be the sort of questions Mike is holding out for – hypothetical voting intention questions with alternative Labour leaders. The only poll to include hypothetical match up questions so far is YouGov’s for the Daily Telegraph at the end of July which showed Miliband doing marginally worse than Brown. Now, strictly speaking even these figures weren’t comparable. A poll asking how people will vote with David Miliband as Labour leader by definition includes David Miliband as a prompt in the question, wheras normal voting intention figures do not include Gordon Brown’s name in the question. A really comparative question should prompt with the name of the party leader in both questions. Even then, people are not necessarily good judges of how they will react to future events. Even if we did get good, solid comparable questions, they wouldn’t be that useful.

Polls before Gordon Brown became Prime Minister used to show a larger Conservative lead than under Tony Blair and now there is a big Conservative lead, lots of people have concluded these questions are useful. It ain’t necessarily so. In different ways those polls were both right and very, very wrong. They were wrong in the sense that Gordon Brown becoming Labour leader actually transformed the position of the political parties and put the Labour party into a strong lead, it was accompanied by lots of positive press coverage, eye-catching first moves like Brown bringing in his ‘G.O.A.T’s’ and a general willingness to give Gordon Brown a chance. It wasn’t too hard to predict that Brown would get this sort of honeymoon, but you wouldn’t have guessed it from the polls that preceeded it and – if Gordon Brown had gone to the country at the beginning of that honeymoon period he may have been able to win an election on the back of that goodwill. Then we really would have been cursing all those pre-Brown hypothetical polls.

In the longer term though people were right about Gordon Brown having worse ratings than Tony Blair. My guess is that those poor ratings before Brown became leader were a reflection of people’s negative reactions toward’s Brown personality. It certainly wasn’t because they thought him weak, indecisive or incompetent, however poor his ratings on those fronts are now: back then Brown was still very highly rated in terms of competence, efficiency and strength. People must have been reacting negatively to his personality and my view is that a lot of negative feelings towards Brown now are for the same reason.

So, how much trust should we place in any hypothetical questions about how people would vote with David Miliband as leader? Well, firstly we should not place any trust at all in them predicting what the short term effect of a change of leader would be. A second chance of leader may or may not produce the same sort of boost in Labour’s ratings as Gordon Brown’s accession did. Some of the media coverage of Gordon Brown in Summer 2007 bordered on hagiography, I’d like to think that in hindsight the media realised they went rather over the top and would be more controlled in their reactions to a second new Labour leader, but never the less, I would expect a strong boost in Labour’s ratings were Gordon Brown to actually go. His net approval rating stands at minus 55 in YouGov’s latest poll, minus 47 in Ipsos MORI’s. The removal of such a strongly negative factor should improve Labour’s ratings unless replaced by someone similarly repellent to voters and in nearly all cases people will give some benefit of the doubt to even an unsympathetic replacement. How much of an initial boost a new Labour leader gets may be the important question if they go for an immediate election, not what the longer term picture would be.

Moving beyond that, can any question tell us how people will react to David Miliband in the longer term as it did with Gordon Brown? Obviously we don’t know for sure. It is possible they are showing something meaningful – perhaps people have decided that they don’t like David Miliband either – too young or too geeky – and he’d be no improvement. Perhaps people have decided that Labour should definitely go regardless of who their leader is, and therefore they will vote the same regardless of who leads them. In either of these hypotheses figures predicting Labour would do just as badly would be a good guide, but I don’t think either are right.

Hypothetical polls now are not likely to be the result of a negative reaction to Miliband himself, or at least, not an accurate one, since people don’t know him well enough to make a judgement. In an ICM poll earlier this month only 53% of people said they would even recognise Miliband, and it seems unlikely that all that 53% have a good idea of what Miliband is like and what he stands for. When people reacted negatively to Gordon Brown’s personality they had 10 years experience of him as a senior minister. In the case of David Miliband they have just over 1. This lack of recognition for Miliband is the reason Nick Sparrow of ICM gives for why ICM haven’t run any questions of this sort. In the second case, it is inconcievable that every swing voter in the country decides their vote purely on government performance. To some degree or other things like leadership, party image, party priorities and policies will all be major factors and David Miliband as leader would change all these things to a degree: people might not believe he would now, but he may yet surprise them.

While I don’t think hypothetical polls of this type would be particularly useful when the alternative leader is not particularly well known to the public, they would nevertheless be important. If these polls did show Miliband doing better they would bolster the courage of people within Labour considering whether to move against Brown. If not they would act against a real Miliband bandwagon from building up in the press (the same thing was evident before Tony Blair stood down – media speculation over a would be alternative would start building up, newspapers would start running lovely soft-focus stories about Alan Johnson, David Miliband or whoever, then polls would suggest that they would be just as bad as Brown and the media would drop them and move on to the next would be alternative!). Given the combination of being not very enlightening, but potentially having great influence, perhaps its not a bad thing no one has been commissioning them!

Three holiday polls

No sooner do I go on holiday for a week than we get three polls in a row.

ICM/Guardian (15-17th August): CON 44%(-1) LAB 29%(nc) LDEM 19%(+3)
YouGov/Sunday Times (14-15th August): CON 45%(-1) LAB 25%(-1) LDEM 18%(+1)
Ipsos MORI (15-17th August): CON 48%(+1) LAB 24%(-3) LDEM 17%(+2)

Polls in August are normally treated with slight scepticism because it’s the holiday season – at any point a certain proportion of people are likely to be away from home on holiday and this has the potential to result in skewed samples if the sort of person who goes on holiday in August is more likely to have a certain sort of view. In fact some pollsters do attempt to control for this by weighting according to the number of foreign holidays people take each year. Presumably raw samples in August find fewer people who take one or more foreign holidays each year than they should, and those people are weighted upwards to correct for any skew this causes.

August is also a period without any real political news beyond unexpected events like those in Georgia. There is little in the way of policy announcements or the day-to-day cut and thrust of politics. A lot of the normal polling figures we see may well be a reflection of good or bad news at the time – that day, that week. In August, when there isn’t any news to speak of, it’s easy to imagine that we might get figures that better reflect the underlying position. For that reason I normally expect a slight improvement for an embattled government over August. It doesn’t always happen, and clearly it hasn’t this year. The government’s position away from constant negative news stories seems to be pretty much the same as it was during them.

These three polls essentially show a stable position – the Conservaties are still in the mid to high forties, with a lead of 20+ points with YouGov, MORI and ComRes or 15+ points with ICM and Populus. The difference, incidentally, is at least partially because of ICM and Populus’s reallocation of those who say don’t know to the party they voted for in 2005. In Populus’s last poll this reduced the Tory lead by 2 points, in ICM’s poll for the Sunday Express earlier this month by 3 points.

Notably the couple of recent polls that showed a fall in Lib Dem support have reversed. In the absence of any reason to expect a sudden fall and recovery in Lib Dem support, it looks that that was just a blip.

Looking at other questions in the polls, both ICM and YouGov had some questions about David Miliband – I’ll come to them in a later post. On other topics, YouGov in the Sunday Times asked about the situation in Georgia. 32% thought the West should have done more to help Georgia, 27% disagreed. 37% thought that Russia was trying to recreate the Soviet Empire and 48% that we should fear them.

YouGov also covered the subject of the NHS. A large majority (73%) of people rejected the idea that NHS treatments should be limited according to cost, thinking that any treatments that exist should be available on the NHS. However asked if they would be willing to pay for them, 46% said they would not be willing to pay extra taxes to make more expensive treatment available. Only 32% agreed. This is not necessarily a direct contradiction – I expect people who wanted all treatments available on the NHS, but were not prepared to pay for them expected other people to pay higher taxes to fund it. It is a notable turnaround however, if you go back three or four years then any question asking if people would be prepared to pay higher taxes to spend on the NHS would have resulted in a resounding yes. It was questionable whether those answers were realistic, or simply people giving the ‘socially acceptable’ answer, but either way there has been a significant shift in attitudes. Asked if people should be able to “top up” their NHS treatment by paying for treatments that the NHS does not fund, 65% thought people should with only 23% opposed.



As you may have guessed, i’m currently on holiday, in the wilds of norfolk where even broadband toggles fear to tread (or at least fear to work). Please use this thread to discuss any exciting polls in my absence – like that ICM one in this mornings Guardian.

The SNP continue to drip-drip the findings of their YouGov poll into the public arena – today’s press release has the constituency voting intentions for the Scottish Parliament, and show a daunting SNP lead over Labour. Topline voting intentions are CON 13%, LAB 25%, LDEM 14%, SNP 44% – a nineteen point lead for the SNP.

If repeated at the next Scottish elections this would produce a stunning landslide for the SNP on the constituency seats. On a uniform swing they would take 57 seats to Labour’s 8, with 6 for the Liberal Democrats and 2 for the Conservatives. Of course, they haven’t released the data on how people would cast their second vote were Labour would pick up a large number of seats due to SNP domination of the constituency MSPs, so we can’t project an overall result based on these figures.

The full tables for the News of the World’s YouGov poll are now available here.

On whether replacing Gordon Brown as leader would make people more or less likely to vote Labour, 21% said it would make them more likely, with 7% saying less likely. While I’m still no great fan of questions like this, YouGov do word them the best way, giving people options of saying they will definitely or definitely not vote Labour anyway, taking most of those people whose vote wouldn’t really change anyway out of picture. 19% of current Tory voters say they would be more likely to vote Labour if Brown went, 25% of Lib Dem voters say they would be more likely to voter Labour without him. Only 11% of current Labour voters would be less likely to vote Labour without Brown. Of course, this registers only negative feeling towards Brown – a replacement leader would bring their own positive or negative responses from the public.

Interestingly – and I don’t think I’ve actually seen this in a survey before – YouGov also asked why people said they would be more likely to vote Labour without Brown. The most popular answers were the rather unenlightening “not a good leader” (68%), “lack of charisma” (52%), “out of touch” (50%), “handling of the economy” (48%), “wasn’t elected” (46%). Other policy issues like crime, health, education were not major concerns.

YouGov asked people who would be the best leader of the Labour party, and included Brown himself in the list – most recent questions have asked only about potential replacements. Despite his relative anonymity, David Miliband beat Gordon Brown by 12% to 10%, though “none of the above” and “don’t know” were easy winners. Amongst Labour voters 32% chose Brown over Miliband (15%).